Comparison of Ethinyl-Estradiol Plus Cyproterone Acetate Versus Metformin Effects on Classic Metabolic Cardiovascular Risk Facto

Home , Troglitazone

Comparison of Ethinyl-Estradiol Plus Cyproterone Acetate Versus Metformin Effects on Classic Metabolic Cardiovascular Risk Factors in Women with the Polycystic Ovary Syndrome

Manuel Luque-Ramı´rez, Francisco A´ lvarez-Blasco, Jose´ I. Botella-Carretero, Elena Martıńez-Bermejo, Miguel A. Lasuncioń, and Hećtor F. Escobar-Morreale Downloaded from https://academic.oup.com/jcem/article/92/7/2453/2598116 by guest on 02 October 2021 Departments of Endocrinology (M.L.-R., F.A.-B., J.I.B.-C., E.M.-B., H.F.E.-M.) and Biochemistry-Research (M.A.L.), Hospital Universitario Ramoń y Cajal and Universidad de Alcala´, E-28034 Madrid, Spain; and Centro de Investigacioń Biome´dica en Red (CIBER) Fisiopatologia Obesidad y Nutricioń (CB06/03) (M.A.L.), Instituto de Salud Carlos III, E-28029 Madrid, Spain

Context: Oral contraceptives may worsen the metabolic profile of Results: Diane35 Diario resulted in higher reductions in hirsutism patients with polycystic ovary syndrome (PCOS), favoring the use of score and serum androgen levels compared with metformin. Men- insulin sensitizers in these patients. strual regularity was restored in all the patients treated with Diane35 Diario compared with only 50% of those receiving met- Objective: The aim of the study was to compare the effects of a formin. Plasma apolipoprotein A-I and HDL-phospholipid levels contraceptive pill on metabolic classic cardiovascular risk factors with increased with Diane35 Diario, whereas metformin did not induce those of the insulin sensitizer metformin. any change in the lipid profile. On the contrary, the insulin sen- Design: We conducted a randomized, parallel, open-label clinical trial. sitivity index increased with metformin but did not change with Diane35 Diario. No differences in the frequencies of abnormalities Setting: The study was conducted at an academic hospital. of glucose tolerance and dyslipidemia were found between both treatments. Patients: Thirty-four consecutive PCOS patients were studied. 35 Interventions: Patients were randomized to oral treatment with Conclusions: Diane Diario appears to be superior to metformin for metformin (850 mg twice daily) or with the Diane35 Diario pill (35 ␮g the control of hyperandrogenism and for the restoration of menstrual of ethinyl-estradiol plus 2 mg of cyproterone acetate) for 24 wk. regularity in PCOS patients, and it is not associated with any clinically relevant worsening in the classic metabolic cardiovascular risk Main Outcome Measures: Hyperandrogenism, lipid profiles, and profile of these women. (J Clin Endocrinol Metab 92: 2453–2461, indexes of glucose tolerance and insulin sensitivity were measured at 2007) baseline and after 12 and 24 wk of treatment.

HE POLYCYSTIC OVARY syndrome (PCOS) is a com- still lacking (9). PCOS patients present increased frequencies T mon endocrine disorder of premenopausal women (1, of glucose intolerance (up to 30–35%) and type 2 diabetes (7 2), characterized by the presence of hyperandrogenism to- to 10%) (10, 11), gestational diabetes (12, 13), an unfavorable gether with ovulatory dysfunction and/or polycystic ovar- lipid profile characterized by decreased high-density li- ian morphology (3). poprotein (HDL)-cholesterol and increased triglycerides lev- Although the primary defect in PCOS appears to be an els (14, 15), and even increased low-density lipoprotein exaggerated androgen synthesis and secretion (4), particu- (LDL)-cholesterol (16, 17) and an increased risk for hyper- larly by ovarian theca cells (5), insulin resistance and obesity tension (18). In conceptual agreement, the prevalence of the may act as triggers of this primary defect (6), explaining the metabolic syndrome is increased in PCOS patients (19, 20), frequent association of PCOS with obesity (6, 7) and with although this finding may result from the frequent associa- insulin resistance (8). tion with obesity (7). Cardiovascular risk factors cluster in patients with PCOS In recent years, the clustering of cardiovascular risk factors and their families, although the demonstration of an increase in PCOS patients raised concern about the optimal drug in cardiovascular morbidity or mortality in these patients is therapy for this disorder. Aside from essential lifestyle rec- ommendations and diet in women with weight excess (21) First Published Online April 10, 2007 and from therapeutic strategies directed toward restoration Abbreviations: Apo, Apolipoprotein; AUC, area(s) under the curve; of fertility, the chronic pharmacological treatment of PCOS BMI, body mass index; HDL, high-density lipoprotein; LDL, low-density is based on the use of two families of drugs, namely oral lipoprotein; oGTT, oral glucose tolerance test; PCOS, polycystic ovary syndrome. contraceptives containing a progestin of low androgenicity or even antiandrogenic properties, or insulin sensitizers. JCEM is published monthly by The Endocrine Society (http://www. endo-society.org), the foremost professional society serving the en- Oral contraceptives have been the mainstay of PCOS phar- docrine community. macological therapy for decades (22). However, in nonhy-

2453 2454 J Clin Endocrinol Metab, July 2007, 92(7):2453–2461 Luque-Ramı´rez et al. • Ethinyl-Estradiol Plus Cyproterone vs. Metformin in PCOS perandrogenic women oral contraceptives might adversely Main outcome measurements influence insulin resistance and glucose tolerance, raising A single investigator (M.L.-R.) was responsible for clinical, anthro- concern about a possible worsening of the unfavorable met- pometric, and physical evaluations at baseline and after 12 and 24 wk abolic cardiovascular risk profile of PCOS patients and fa- of treatment. Clinical and anthropometrical variables included the voring the use of the metabolically safer insulin-sensitizer above-mentioned hirsutism score, body mass index (BMI), waist circumference, and waist to-hip ratio. The latter was calculated by dividing drugs (23). the minimal waist circumference by the hip circumference at the level Yet at present there is not enough scientifically sound of greater trochanters, using a nonstretchable measuring tape. The per- evidence supporting this recommendation (24). The present centage of body fat with respect to total body weight was estimated study summarizes the results of a randomized clinical trial using a body fat monitor (Omron BF 300; Omron Corp., Kyoto, Japan). Serum and plasma samples were obtained between d 5 and 10 of a comparing the effects of an antiandrogenic low-dose oral spontaneous menstrual bleeding or after excluding pregnancy in am- contraceptive pill with those of the insulin sensitizer met- enorrheic patients. After a 3-d 300-g carbohydrate diet and 12-h over- formin on the classic metabolic cardiovascular risk factors of night fasting, basal samples were obtained for measurement of an an-

PCOS patients. drogenic profile consisting in serum free testosterone, androstenedione, Downloaded from https://academic.oup.com/jcem/article/92/7/2453/2598116 by guest on 02 October 2021 and dehydroepiandrosterone sulfate concentrations (3). Then, a 75-g oral glucose tolerance test (oGTT) was performed, and samples were ob- Patients and Methods tained for measurement of serum insulin and plasma glucose at 0, 30, 60, 90, and 120 min. Samples were immediately centrifuged, and serum Patients was separated and frozen at Ϫ 30 C until assayed. The technical characteristics of the assays employed for plasma glu- The present study is the first report of a broader project aiming to cose and serum hormone measurements have been described elsewhere compare systematically the effects of an oral contraceptive and met- (28, 29). Free testosterone levels were calculated from total testosterone formin on classic and nonclassic cardiovascular risk factors and on and SHBG concentrations (30). The composite insulin sensitivity index several indexes of cardiovascular function and performance, by using a was calculated from the circulating glucose and insulin concentrations randomized, open-label, parallel clinical trial comparing both treat- during the oGTT (31). The areas under the curve (AUC) for glucose and ments (www.clinicaltrials.gov; NLM Identifier NCT00428311). insulin during the oGTT were determined according to the mathematical Thirty-four consecutive PCOS patients were recruited. The diagnosis method described by Tai (32). of PCOS was based on the presence of clinical and/or biochemical Disorders of glucose tolerance were defined from the circulating hyperandrogenism, oligoovulation, and exclusion of secondary etiolo- glucose concentrations at 0 and 120 min during oGTT according to the gies (3, 25). Hirsutism was defined by a modified Ferriman-Gallwey criteria of the American Diabetes Association (33). Circulating HDL- score above 7 (26) and oligomenorrhea (more than six cycles longer than cholesterol and phospholipid levels were measured by enzymatic meth- 36 d in the previous year) or amenorrhea (absence of menstruation for ods after precipitation of plasma with phosphotungstic acid and Mg2ϩ 3 consecutive months) (27), or luteal phase progesterone measurements (Roche Molecular Biochemicals GmbH, Mannheim, Germany). Total less than 4 ng/ml (12.72 nmol/liter) in women with regular menstrual cholesterol and triglyceride levels were determined by enzymatic meth- cycles were considered indicative of oligoovulation. Secondary etiolo- ods (Menarini Diagnostica, Florence, Italy). LDL-cholesterol concentra- gies, including hyperprolactinemia, thyroid dysfunction, Cushing’s syn- tions were estimated by Friedewald’s equation (34). Circulating apoli- drome, congenital adrenal hyperplasia, and virilizing tumors, were ac- poprotein (Apo) AI, Apo B100, and lipoprotein (a) levels were tively ruled out in all the patients. determined by kinetic immunonephelometry (Dade Behring, Deerfield, None of the patients had a personal history of hypertension, diabetes IL). The diagnosis of dyslipidemia was established according to the mellitus, or cardiovascular events, or received treatment with oral con- criteria of the National Cholesterol Education Program Adult Treatment traceptives, antiandrogens, insulin sensitizers, or drugs that might in- Panel III guidelines (35). terfere with blood pressure regulation, lipid profile, or carbohydrate metabolism for the previous 6 months. Written informed consent was Statistical analysis obtained from all the participants or their legal representatives, and the study was approved by the local ethics committee and by the Spanish Data are shown as means Ϯ sd and raw numbers (percentages) unless Agency of Medicines. otherwise stated. The differences in frequencies among groups were analyzed by ␹2 or Fisher’s exact tests as appropriate. The changes in the frequencies of metabolic disorders during the study were analyzed by Interventions McNemar’s test, applying a Bonferroni correction to the level of significance to correct for multiple comparisons. After providing informed consent, patients were randomized to re- For continuous variables, normality was assessed by the Kolmog- ceive an antiandrogenic low-dose oral contraceptive pill in cycles of 28 d orov-Smirnov test, and logarithmic or square root transformations were (21 pills containing 35 ␮g of ethinyl-estradiol plus 2 mg of cyproterone 35 applied as needed to ensure normality. The baseline characteristics of the acetate followed by 7 placebo pills, Diane Diario; Schering Espan˜a S.A., patients randomized to receive Diane35 Diario or metformin were com- Madrid, Spain), or 850 mg of metformin (Dianben; Merck Farma y pared by unpaired t test. Data were submitted to a repeated-measures Quı´mica S.A., Mollet del Valle´s, Barcelona, Spain) twice daily for 24 wk. general lineal model including the arm of treatment as the between- Simple randomization was conducted using blocks of 10 sealed opaque 35 subjects effect, and the visit (baseline, 12 and 24 wk) as the within- envelopes assigning five patients to receive Diane Diario and five subjects effect. To evaluate the differences in the response to each treat- patients to receive metformin. One investigator generated the random- ment, the interaction between the between-subjects and within-subjects ization envelopes, whereas another enrolled and assigned the partici- effect was calculated. pants to their arm of treatment. No masking method was used after A priori sample size analysis was performed using the online calcu- randomization. lator provided by the Massachusetts General Hospital Mallinckrodt Treatment was started the first day of a spontaneous menstrual cycle General Clinical Research Center (http://hedwig.mgh.harvard.edu/ or, in women with amenorrhea, after excluding pregnancy by proper sample_size/size.html). Because the possible advantages of metformin testing. Metformin was started at a 425-mg dose twice daily during the over Diane35 Diario rely on the presumed opposite effects of both drugs first week of treatment with the aim of minimizing gastrointestinal side on insulin resistance, we used the difference in surrogate markers of effects. Women randomized to metformin were advised to use barrier insulin resistance as the primary outcome for sample size analysis. An contraception throughout the study. All the patients were instructed to experimental design including a total of 22 patients would have power maintain a diet containing 25–30 kcal per kg of body weight per day and above 0.80 to detect differences in fasting insulin levels between treat- moderate physical activity throughout the trial, although these measures ments such as those previously reported using a similar experimental were not stressed thereafter. design (36, 37). Luque-Ramı´rez et al. • Ethinyl-Estradiol Plus Cyproterone vs. Metformin in PCOS J Clin Endocrinol Metab, July 2007, 92(7):2453–2461 2455

Because seven patients discontinued metformin for different reasons resulted mostly from the decrease observed in the patients (Fig. 1), results obtained when considering only the patients completing treated with Diane35 Diario, which was much more marked the three visits of the protocol were also confirmed by intention-to-treat than those observed with metformin (Fig. 2). Similarly, only analysis assuming, for patients who did not complete the study, that the 35 dependent variables had not changed at the missing visits with respect treatment with Diane Diario induced a decrease in serum to the values observed in the previous visit. P Ͻ 0.05 was considered dehydroepiandrosterone sulfate levels (Fig. 2). Finally, at the statistically significant. Other than sample size calculations, analyses end of the study, menstrual regularity was restored as ex- were performed using SPSS 10.0 for Macintosh (SPSS Inc., Chicago, IL). pected in all the patients on Diane35 Diario but only in six patients (50%) on metformin. Results Recruitment and flow of the study Effects of treatments on the lipid profile The study was conducted from April 2004 to December 2006. Thirty-four PCOS patients (age 24 Ϯ 6 yr; BMI 30.0 Ϯ As occurred with clinical and biochemical hyperandro- 2 2 genism, the changes observed in the lipid profile were Ͻ Downloaded from https://academic.oup.com/jcem/article/92/7/2453/2598116 by guest on 02 October 2021 6.3 kg/m ; nine had a BMI 25 kg/m , 10 had a BMI 25–29.9 35 kg/m2, and 15 had a BMI Ն 30 kg/m2) agreed to participate mainly related to the effects of Diane Diario. In the women and, after giving written informed consent, were random- treated with this drug, plasma Apo A-I and HDL-phospho- ized to receive Diane35 Diario or metformin. The baseline lipid concentrations increased significantly, explaining the characteristics of each group are summarized in Table 1, increase observed in these lipids, and in plasma HDL-cho- showing no differences among the patients randomized to lesterol concentrations, in the whole group of patients when each drug. All the patients treated with Diane35 Diario com- compared with baseline values (Fig. 3). On the contrary, no pleted the study, whereas seven of the 19 patients assigned statistically significant changes were observed in plasma to- to metformin (37%) did not complete the study for the rea- tal and LDL-cholesterol, triglycerides, Apo B100, and li- sons detailed in Fig. 1. Of those, mild to moderate gastro- poprotein (a) levels compared with baseline concentrations intestinal side effects of metformin were responsible for the (Fig. 3). dropout in two women, whereas Diane35 Diario was well tolerated by all the women. Effects of treatments on indexes of glucose tolerance and insulin sensitivity Effects of treatments on clinical characteristics and serum Indexes of insulin sensitivity improved in the whole series androgen profile of PCOS patients, as the insulin sensitivity index increased, There were no differences with respect to baseline in BMI, and fasting insulin and the insulin AUC during the oGTT waist circumference, waist to hip ratio, and fat mass ex- decreased, compared with baseline (Fig. 4). Of these changes, pressed as percentage of body weight (Fig. 2). When con- only the increase in the insulin sensitivity index was specif- sidering PCOS patients as a whole, the hirsutism score and ically related to the metformin arm (Fig. 4), as suggested by serum free testosterone and androstenedione levels de- a statistically significant interaction in the general linear creased with treatment (Fig. 2), yet detailed analysis of the model in the analysis of patients completing the study (al- interaction between the within- and between-subjects effects though this particular interaction was not confirmed by the of the general linear model revealed that these differences intention-to-treat analysis). On the contrary, fasting glucose

PCOS patients recruited (n = 34)

Baseline evaluation Simple randomization (Blocks of 10 sealed envelopes, assigning 5 patients to each arm of treatment)

Diane35 Diario Metformin (n = 15) (n = 19) FIG. 1. Flow of the study. Protocol violation (n = 2) Gastro-intestinal side effects (n = 2) Pregnancy (n = 1)

(n = 15) Evaluation (n = 14) (12 weeks)

Protocol violation (n = 1) Lost to follow-up (n = 1)

(n = 15) Evaluation (n = 12) (24 weeks) End of study 2456 J Clin Endocrinol Metab, July 2007, 92(7):2453–2461 Luque-Ramı´rez et al. • Ethinyl-Estradiol Plus Cyproterone vs. Metformin in PCOS

TABLE 1. Baseline characteristics of the patients randomized to receive Diane35 Diario or metformin

Diane35 Diario Metformin P value (n ϭ 15) (n ϭ 19) Age (yr) 23.4 Ϯ 5.6 25.1 Ϯ 6.6 0.445 Smokers, n (%) 6 (40) 8 (42) 0.901 BMI (kg/m2) 29.2 Ϯ 5.7 30.5 Ϯ 6.9 0.563 Waist circumference (cm) 83 Ϯ 12 89 Ϯ 18 0.306 Waist to hip ratio 0.79 Ϯ 0.06 0.82 Ϯ 0.11 0.292 Fat mass (% of body weight) 33 Ϯ 634Ϯ 8 0.759 Hirsutism score 11 Ϯ 510Ϯ 6 0.629 Free testosterone (ng/dl) 1.1 Ϯ 0.4 1.3 Ϯ 0.6 0.210 Androstenedione (ng/ml) 3.5 Ϯ 0.8 3.9 Ϯ 1.1 0.226 Dehydroepiandrosterone sulfate (ng/ml) 2,738 Ϯ 1,022 2,250 Ϯ 933 0.156 Total cholesterol (mg/dl) 154 Ϯ 20 159 Ϯ 28 0.529 HDL-cholesterol (mg/dl) 39 Ϯ 11 41 Ϯ 11 0.536 Downloaded from https://academic.oup.com/jcem/article/92/7/2453/2598116 by guest on 02 October 2021 HDL-phospholipids (mg/dl) 100 Ϯ 18 108 Ϯ 17 0.244 LDL-cholesterol (mg/dl) 95 Ϯ 19 101 Ϯ 28 0.459 Triglycerides (mg/dl) 106 Ϯ 96 89 Ϯ 45 0.548 Apo A-I (mg/dl) 125 Ϯ 18 129 Ϯ 18 0.556 Apo B100 (mg/dl) 76 Ϯ 22 83 Ϯ 20 0.376 Lipoprotein (a) (mg/dl) 14 Ϯ 23 29 Ϯ 33 0.132 Fasting glucose (mg/dl) 93 Ϯ 893Ϯ 8 0.993 Fasting insulin (␮IU/ml) 14 Ϯ 11 18 Ϯ 16 0.429 AUC glucose (mg/dl⅐120 min) 15,143 Ϯ 2,263 15,982 Ϯ 4,080 0.529 AUC insulin (␮IU/ml⅐120 min) 9,923 Ϯ 5,793 10,158 Ϯ 5,684 0.906 Insulin sensitivity index 4.4 Ϯ 3.5 3.8 Ϯ 2.4 0.564

Data are expressed as means Ϯ SD, or raw numbers (percent). Data were submitted to unpaired t test or to ␹2 test, as appropriate. To convert to SI units, multiply free testosterone by 34.67 (result in pmol/liter), androstenedione by 3.49 (result in nmol/liter), dehydroepiandrosterone sulfate by 0.002714 (result in ␮mol/liter), cholesterol by 0.0259 (result in mmol/liter), triglycerides by 0.0113 (result in mmol/liter), glucose by 0.0555 (result in mmol/liter) and insulin by 6.945 (result in pmol/liter). AUC, Area under the curve during the oGTT. and the glucose AUC during the oGTT showed no changes tolerance and in two of the women presenting with impaired with respect to baseline values in the whole series of PCOS fasting glucose, one patient presenting with normal glucose patients, although fasting glucose increased slightly in the tolerance at baseline developed impaired fasting glucose and patients treated with Diane35 Diario at the end of the study, another woman developed impaired glucose tolerance after compared with the evaluations performed at baseline and this treatment. after 12 wk of treatment (Fig. 4). Discussion Frequencies of disorders of glucose tolerance and Our present results confirm that both Diane35 Diario and dyslipidemia at baseline and throughout the study metformin are safe and effective in the treatment of PCOS There were no statistically significant differences in the patients, both drugs showing overall beneficial effects on frequencies of disorders of glucose tolerance and dyslipide- most of the clinical complaints characteristic of this com- mia between the arms of treatment, or at baseline and after mon disorder. However, our data also suggest that Diane35 12 or 24 wk of treatment either when considering PCOS Diario is a more efficient way of treating hyperandro- patients as a whole or when considering each arm of treat- genism and menstrual dysfunction, considering that the ment separately (Table 2). improvement in hirsutism, the amelioration of hyperan- Of note, both the increase in patients with impaired fasting drogenemia, and the restoration of regular menstrual cy- glucose and the decrease in the number of patients with low cles occur earlier and more frequently with Diane35 Diario HDL-cholesterol levels after treatment with Diane35 Diario than with metformin. were far from reaching statistical significance. Also, it should On the other hand, although metformin clearly outper- be noted that in the two patients who developed impaired forms Diane35 Diario in improving insulin resistance, our fasting glucose after treatment with Diane35 Diario, fasting present data indicate that Diane35 Diario is a safe drug when glucose was minimally increased, i.e. to 100 mg/dl and 101 considering classic metabolic cardiovascular risk factors, and mg/dl (5.6 mmol/liter) respectively, and there was another might even have favorable effects on the lipid profile of patient who presented with impaired fasting glucose at base- PCOS patients. To this regard, whereas the improvement in line that returned to normal fasting glucose levels after treat- insulin resistance found with metformin was not related to ment with the oral contraceptive. Similarly, the changes in any change in the lipid profile of PCOS patients, Diane35 glucose tolerance in the patients treated with metformin did Diario actually induced an increase in plasma Apo A-I and not follow a discernible pattern despite the improvement in HDL-phospholipid levels, explaining the increase in plasma insulin resistance observed in these women as a group. HDL-cholesterol concentrations observed in the whole Whereas normalization of glucose tolerance after 24 wk of group of PCOS patients. This finding is especially important treatment with metformin was observed in the patient with considering that decreased HDL-cholesterol levels are type 2 diabetes, in another patient with impaired glucose among the commonest lipid abnormalities associated with ramn,bt yaayi fptet h opee h td n yitnint-ra analysis. intention-to-treat by and study the completed who patients of analysis by both treatment, ftetet nyi h neto-otetaayi.‡, analysis. intention-to-treat the in only treatment, of 9,ytasmn o ainswoddntcmlt h td httedpnetvralsddntcag ttemsigvst ihrespect with visits missing the at †, change treatment. not of did variables arm dependent the *, the of Dehydroepiandrosterone-sulfate. that DHEAS, irrespective study visit. patients, previous the of the complete group in not observed did values who the patients to for (n assuming study yet the 19), completing patients the of F Luque-Ramı efri o 4w.Dt r xrse smeans as expressed are Data wk. 24 for metformin IG .Cagsi lnclcaatrsisadsrmadoe rflso CSptet umte otetetwt Diane with treatment to submitted patients PCOS of profiles androgen serum and characteristics clinical in Changes 2. . ´rez

Androstenedione Hirsutism score Waist-to-hip ratio Body mass index al et

(ng/mL) (kg/m2) Cyproterone Plus Ethinyl-Estradiol • . 0.6 0.7 0.8 0.9 1.0 20 25 30 35 40 12 16 1 2 3 4 5 0 4 8 01224 ϭ 2 n loa neto-otetaayi nldn l h ainstetdwt efri (n metformin with treated patients the all including analysis intention-to-treat an also and 12) ‡‡ *‡ *‡ *† Ϯ Metformin (intention-to-treat analysis, n = 19) = n analysis, (intention-to-treat Metformin 12) = n study, the completing (patients Metformin Diane SEM vs efri nPO lnEdcio ea,Jl 07 92(7):2453–2461 2007, July Metab, Endocrinol Clin J PCOS in Metformin . P eas ftelredootrt ntemtomnam h iue hwteanalysis the show figures the arm, metformin the in rate dropout large the of Because . 35 Ͻ P Diario (n (n = Diario 15) Weeks oftreatment .5frtedfeecsi h hne fec aibedpnigo h r of arm the on depending variable each of changes the in differences the for 0.05 Ͻ .5frtedfeecsi h hne fec aibedpnigo h arm the on depending variable each of changes the in differences the for 0.05

DHEAS Free testosterone Fat mass Waist circumference (ng/mL) (ng/dL) (% of body weight) (cm) 1000 2000 3000 4000 5000 100 110 0.0 0.5 1.0 1.5 2.0 70 80 90 25 30 35 40 45 P 01224 Ͻ .5cmae ihbsln ausi h whole the in values baseline with compared 0.05 35 iroor Diario 2457

ϭ Downloaded from https://academic.oup.com/jcem/article/92/7/2453/2598116 by guest on 02 October 2021 October 02 on guest by https://academic.oup.com/jcem/article/92/7/2453/2598116 from Downloaded 2458 J Clin Endocrinol Metab, July 2007, 92(7):2453–2461 Luque-Ramı´rez et al. • Ethinyl-Estradiol Plus Cyproterone vs. Metformin in PCOS

Diane35 Diario (n = 15) Metformin (patients completing the study, n = 12) Metformin (intention-to-treat analysis, n = 19)

200 140

180 120

160 100 mg/dL) (mg/dL) (

140 80 Downloaded from https://academic.oup.com/jcem/article/92/7/2453/2598116 by guest on 02 October 2021 Total cholesterol Total LDL - cholesterol - LDL

120 60 80 160

**†140 60

120 (mg/dL) 40 (mg/dL) 100 HDL - cholesterol HDL - phospholipid 20 80

200 200

150 175 *†

100 150 (mg/dL) (mg/dL) Apo A-I 50 125 Triglycerides

0 100

120 100

80 90 60

40 (mg/dL) 60 (mg/dL) Apo B100 Apo

Lipoprotein (a) Lipoprotein 20

30 0 01224 01224

Weeks of treatment

FIG. 3. Changes in the lipid profile of PCOS patients submitted to treatment with Diane35 Diario or metformin for 24 wk. Data are expressed as means Ϯ SEM. Because of the large dropout rate in the metformin arm, the figures show the analysis of the patients completing the study (n ϭ 12) and also an intention-to-treat analysis including all the patients treated with metformin (n ϭ 19), yet assuming for patients who did not complete the study that the dependent variables did not change at the missing visits with respect to the values observed in the previous visit. *, P Ͻ 0.05 compared with baseline values in the whole group of patients, irrespective of the arm of treatment. †, P Ͻ 0.05 for the differences in the changes of each variable depending on the arm of treatment, both by analysis of patients who completed the study and by intention-to-treat analysis.

PCOS (14, 15) and are a recognized cardiovascular risk factor Considering our present finding with those reported by (38). Furthermore, similar results have been reported earlier others, it appears that Diane35 Diario has an effect on the in the few studies actually comparing metformin with the regulation of HDL-cholesterol metabolism. This possibly in- particular oral contraceptive formulation used here (39, 40). volves the activity of plasma hepatic lipase activity, an en- Luque-Ramı´rez et al. • Ethinyl-Estradiol Plus Cyproterone vs. Metformin in PCOS J Clin Endocrinol Metab, July 2007, 92(7):2453–2461 2459

zyme that is remarkably sex steroid sensitive: estrogens de- 35 Diane Diario (n = 15) crease its activity, whereas androgens and androgenic Metformin (patients completing the study, n = 12) progestins increase it (41). In conceptual agreement, and Metformin (intention-to-treat analysis, n = 19) although certain oral contraceptives may worsen the lipid profile in the general population (42, 43), the amelioration of 110 hyperandrogenism in PCOS women, combined with the antiandrogenic properties of cyproterone acetate and the estrogen 35 100 * component of Diane Diario might overcome this undesirable effect, explaining its beneficial effect in PCOS patients (44). 35 90 Moreover, Diane Diario had a minor impact on glucose tolerance because only a minimal increase in fasting glucose, (mg/dL) which was not accompanied by a simultaneous increase in 80

the AUC for plasma glucose during the oGTT, was observed Downloaded from https://academic.oup.com/jcem/article/92/7/2453/2598116 by guest on 02 October 2021 Fasting glucose between wk 12 and 24 of treatment. Furthermore, fasting 70 insulin and the AUC for insulin during the oGTT decreased 40 during treatment irrespective of the drug administered, and the frequencies of abnormalities in glucose tolerance did not 35 30 † change significantly during treatment with Diane Diario. Therefore, our present results are in agreement with previous 35 20 reports showing minor effects, if any, of Diane Diario on

UI/mL) glucose tolerance in PCOS patients (37, 39). Of note, the µ µ µ µ ( 10 amelioration of insulin resistance observed during the 24 wk

Fasting insulin of our study in the patients treated with metformin was not accompanied by an actual reduction in the frequencies of 0 abnormalities in glucose tolerance in these women. 20000 Importantly, treatment with metformin was associated with a high dropout rate in our trial, in the range of the 38% 18000 (39) and 45% (40) of patients treated with this drug in previous clinical trials comparing Diane35 Diario with met- 16000 formin, respectively. Aside from the gastrointestinal side effects characteristic of metformin (45), which were directly 14000 related to the dropout of two of our patients, the usually AUC glucose (mg/dL/120 min) suboptimal response of hirsutism with metformin (46) might 12000 have contributed to the loss of patients in our study considering that hirsutism is a frequent referral complaint in young 20000 PCOS patients such as ours (47). Nevertheless, the confir- mation by intention-to-treat analysis of almost all the results 15000 † obtained analyzing the patients who completed the study rules out a significant impact of the high dropout rate ob- 10000 served in the metformin arm of our trial. On the contrary, tolerance of Diane35 Diario was excellent, and no patient

UI/mL/120 min) UI/mL/120 5000 dropped out of this arm of treatment. AUC insulin µ µ µ µ ( However, our study is not free from limitations, especially 0 the relatively short duration of treatment considering that 12 PCOS is a chronic condition that requires long-term interventions. Yet also, it must be noted that our present results †‡ were obtained in a relatively small sample of PCOS patients 8 dropout rate in the metformin arm, the figures show the analysis of the patients completing the study (n ϭ 12) and also an intention-to- 4 treat analysis including all the patients treated with metformin (n ϭ 19), yet assuming for patients who did not complete the study that the dependent variables did not change at the missing visits with respect to the values observed in the previous visit. AUC, Area under the 0 curve during the oGTT. *, P Ͻ 0.05 for the differences in the changes Insulin sensitivityindex 01224 of each variable depending on the arm of treatment, in both analysis of patients who completed the study and in intention-to-treat anal- Weeks of treatment ysis. †, P Ͻ 0.05 compared with baseline values in the whole group of patients, irrespective of the arm of treatment. ‡, P Ͻ 0.05 for the FIG. 4. Changes in indexes of glucose tolerance and insulin sensi- differences in the changes of each variable depending on the arm of tivity in PCOS patients submitted to treatment with Diane35 Diario treatment, only in patients completing the study but not by intention- or metformin for 24 wk. Data are means Ϯ SEM. Because of the large to-treat analysis. 2460 J Clin Endocrinol Metab, July 2007, 92(7):2453–2461 Luque-Ramı´rez et al. • Ethinyl-Estradiol Plus Cyproterone vs. Metformin in PCOS

TABLE 2. Frequencies of disorders of glucose tolerance and dyslipidemia throughout the study

Metformin 35 ϭ Diane Diario (n 15) Patients completing the study Intention-to-treat analysis (n ϭ 12) (n ϭ 19) Baseline 12 wk 24 wk Baseline 12 wk 24 wk Baseline 12 wk 24 wk Glucose tolerance Normal 10 (67) 12 (80) 8 (53) 10 (83) 9 (75) 10 (83) 14 (74) 14 (74) 15 (79) Impaired fasting glucose 4 (27) 1 (7) 5 (33) 0 (0) 0 (0) 1 (8) 4 (21) 4 (21) 3 (16) Impaired glucose tolerance 2 (13) 2 (13) 2 (13) 1 (8) 3 (25) 1 (8) 3 (16) 5 (26) 3 (16) Type 2 diabetes 0 (0) 0 (0) 0 (0) 1 (8) 0 (0) 0 (0) 1 (5) 0 (0) 0 (0) Global 5 (33) 3 (20) 7 (47) 2 (17) 3 (25) 2 (17) 5 (26) 5 (26) 4 (21) Dyslipidemia Increased total cholesterol 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)

Increased LDL-cholesterol 0 (0) 1 (7) 0 (0) 0 (0) 0 (0) 1 (8) 1 (5) 1 (5) 2 (11) Downloaded from https://academic.oup.com/jcem/article/92/7/2453/2598116 by guest on 02 October 2021 Decreased HDL-cholesterol 8 (53) 4 (27) 4 (27) 4 (33) 6 (50) 4 (33) 9 (47) 11 (58) 9 (47) Hypertriglyceridemia 2 (13) 1 (7) 2 (13) 1 (8) 0 (0) 0 (0) 1 (5) 0 (0) 0 (0) Global 8 (53) 5 (33) 4 (27) 4 (33) 6 (50) 4 (33) 9 (47) 11 (58) 9 (47) Data are expressed as raw numbers (percent). There were no statistically significant differences in the frequencies of metabolic abnormalities between the arms of treatment, nor were there statistically significant changes in these frequencies during treatment when considering all patients as a whole, or each arm of treatment separately. from Spain in whom overweight and obesity were frequent Address all correspondence and requests for reprints to: Hećtor F. findings and who presented with the classic hyperandro- Escobar-Morreale, M.D., Ph.D., Department of Endocrinology, Hospital genic and oligoovulatory PCOS phenotype. Therefore any Universitario Ramoń y Cajal and Universidad de Alcala´, Carretera de Colmenar Viejo Km 9,1. 28034 Madrid, Spain. E-mail: hescobarm.hrc@ extrapolation of these results to PCOS patients of different salud.madrid.org. race, ethnicity, and grade of obesity or presenting with dif- This study was supported by the Spanish Ministry of Health and ferent PCOS phenotypes should be made with prudence. Consumer Affairs, Instituto de Investigacioń Carlos III Grants Fondo de In summary, Diane35 Diario and metformin are safe and Investigacioń Sanitaria PI050341 and PI050551 and Red de Diabetes y Enfermedades Metabo´licas Asociadas (REDIMET) RD06/0015/0007; effective single drugs for the treatment of PCOS. However, 35 Ministry of Education and Science Grant SAF2005-07038; and by eco- in unselected patients presenting with this disorder, Diane nomic aid from Hospital Ramoń y Cajal. Diario appears to be superior in terms of control of clinical Present address for M.L.-R.: Department of Endocrinology, Hospital and biochemical hyperandrogenism and restoration of men- Universitario de La Princesa, Madrid, Spain. strual regularity. Furthermore, this superior efficacy is not Disclosure Information: All authors have nothing to declare. obtained at the expense of any clinically significant worsen- References ing in the metabolic cardiovascular risk profile of these women, yet, much on the contrary, use of Diane35 Diario 1. Asuncion M, Calvo RM, San Millan JL, Sancho J, Avila S, Escobar-Morreale HF 2000 A prospective study of the prevalence of the polycystic ovary syn- might even result in an increase in plasma Apo A-I and drome in unselected Caucasian women from Spain. J Clin Endocrinol Metab HDL-phospholipid levels that is not observed with met- 85:2434–2438 formin. Also, and although only metformin improved insu- 2. Diamanti-Kandarakis E, Christakou C 2006 Prevalence, definition and clinical manifestations of polycystic ovary syndrome. Endocrinol Nutr 53(Suppl 1): lin resistance, this drug did not have a uniform beneficial 25–33 impact on the disorders of glucose tolerance present in PCOS 3. Azziz R, Carmina E, Dewailly D, Diamanti-Kandarakis E, Escobar-Morreale patients that could justify by itself its routine use in all PCOS HF, Futterweit W, Janssen OE, Legro RS, Norman RJ, Taylor AE, Witchel SF 2006 Position statement: criteria for defining polycystic ovary syndrome as a cases. predominantly hyperandrogenic syndrome: An Androgen Excess Society We conclude that Diane35 Diario and metformin should be guideline. J Clin Endocrinol Metab 91:4237–4245 4. Escobar-Morreale HF, Luque-Ramı´rez M, San Millan JL 2005 The molecular- considered for the treatment of PCOS patients, yet the choice for genetic basis of functional hyperandrogenism and polycystic ovary syndrome. each drug requires careful consideration of the patient’s clinical Endocr Rev 26:251–282 characteristics and preference. Furthermore, our present results 5. Nelson VL, Qin K, Rosenfield RL, Wood JR, Penning TM, Legro RS, Strauss 3rd JF, McAllister JM 2001 The biochemical basis for increased testosterone stress that a putative deterioration of the metabolic cardiovas- production in theca cells propagated from patients with polycystic ovary cular risk profile of PCOS patients during administration of syndrome. J Clin Endocrinol Metab 86:5925–5933 Diane35 Diario is not supported by current scientific evidence 6. Gambineri A, Pasquali R 2006 Insulin resistance, obesity and metabolic syndrome in polycystic ovary syndrome. Endocrinol Nutr 53(Suppl 1):49–55 and therefore should not be weighed against the use of this drug 7. Alvarez-Blasco F, Botella-Carretero JI, San Millan JL, Escobar-Morreale HF in any treatment decision making. 2006 Prevalence and characteristics of the polycystic ovary syndrome in overweight and obese women. Arch Intern Med 166:2081–2086 8. Dunaif A 1997 Insulin resistance and the polycystic ovary syndrome: mech- Acknowledgments anism and implications for pathogenesis. Endocr Rev 18:774–800 9. Legro RS 2003 Polycystic ovary syndrome and cardiovascular disease: a pre- We thank Ms. Genoveva Gonza´lez from the Laboratory of the De- mature association? Endocr Rev 24:302–312 partment of Endocrinology of Hospital Ramoń y Cajal for excellent 10. Ehrmann DA, Cavaghan MK, Barnes RB, Imperial J, Rosenfield RL 1999 technical help, and Ms. Francisca Cerrato from the Department of Bio- Prevalence of impaired glucose tolerance and diabetes in women with poly- chemistry-Research for excellent help with the determination of the lipid cystic ovary syndrome. Diabetes Care 22:141–146 profiles. 11. Legro RS, Kunselman AR, Dodson WC, Dunaif A 1999 Prevalence and predictors of risk for type 2 diabetes mellitus and impaired glucose tolerance in polycystic ovary syndrome: a prospective, controlled study in 254 affected Received February 7, 2007. Accepted March 30, 2007. women. J Clin Endocrinol Metab 84:165–169 Luque-Ramı´rez et al. • Ethinyl-Estradiol Plus Cyproterone vs. Metformin in PCOS J Clin Endocrinol Metab, July 2007, 92(7):2453–2461 2461

12. Boomsma CM, Eijkemans MJ, Hughes EG, Visser GH, Fauser BC, Macklon methods for the estimation of free testosterone in serum. J Clin Endocrinol NS 2006 A meta-analysis of pregnancy outcomes in women with polycystic Metab 84:3666–3672 ovary syndrome. Hum Reprod Update 12:673–683 31. Matsuda M, DeFronzo RA 1999 Insulin sensitivity indices obtained from oral 13. Lo JC, Feigenbaum SL, Escobar GJ, Yang J, Crites YM, Ferrara A 2006 glucose tolerance testing: comparison with the euglycemic insulin clamp. Increased prevalence of gestational diabetes mellitus among women with Diabetes Care 22:1462–1470 diagnosed polycystic ovary syndrome: a population-based study. Diabetes 32. Tai MM 1994 A mathematical model for the determination of total area under Care 29:1915–1917 glucose tolerance and other metabolic curves. Diabetes Care 17:152–154 14. Rajkhowa M, Neary RH, Kumpatla P, Game FL, Jones PW, Obhrai MS, 33. The Expert Committee on the Diagnosis and Classification of Diabetes Clayton RN 1997 Altered composition of high density lipoproteins in women Mellitus 2003 Follow-up report on the diagnosis of diabetes mellitus. Diabetes with the polycystic ovary syndrome. J Clin Endocrinol Metab 82:3389–3394 Care 26:3160–3167 15. Berneis K, Rizzo M, Lazzaroni V, Fruzzetti F, Carmina E 2007 Atherogenic 34. Friedewald WT, Levy RI, Fredrickson DS 1972 Estimation of the concentra- lipoprotein phenotype and low-density lipoproteins size and subclasses in tion of low-density lipoprotein cholesterol in plasma, without use of the pre- women with polycystic ovary syndrome. J Clin Endocrinol Metab 92:186–189 parative ultracentrifuge. Clin Chem 18:499–502 16. Talbott E, Clerici A, Berga SL, Kuller L, Guzick D, Detre K, Daniels T, 35. National Cholesterol Education Program (NCEP) Expert Panel on Detection, Engberg RA 1998 Adverse lipid and coronary heart disease risk profiles in Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult young women with polycystic ovary syndrome: results of a case-control study. Treatment Panel III) 2002 Third report of the National Cholesterol Education J Clin Epidemiol 51:415–422 Program (NCEP) Expert Panel on detection, evaluation, and treatment of high 17. Legro RS, Kunselman AR, Dunaif A 2001 Prevalence and predictors of dys- blood cholesterol in adults (Adult Treatment Panel III) final report. Circulation Downloaded from https://academic.oup.com/jcem/article/92/7/2453/2598116 by guest on 02 October 2021 lipidemia in women with polycystic ovary syndrome. Am J Med 111:607–613 106:3143–3421 18. Lo JC, Feigenbaum SL, Yang J, Pressman AR, Selby JV, Go AS 2006 Epi- 36. Morin-Papunen LC, Vauhkonen I, Koivunen RM, Ruokonen A, Marti- demiology and adverse cardiovascular risk profile of diagnosed polycystic kainen HK, Tapanainen JS 2000 Endocrine and metabolic effects of metformin ovary syndrome. J Clin Endocrinol Metab 91:1357–1363 versus ethinyl estradiol-cyproterone acetate in obese women with polycystic 19. Apridonidze T, Essah PA, Iuorno MJ, Nestler JE 2005 Prevalence and char- ovary syndrome: a randomized study. J Clin Endocrinol Metab 85:3161–3168 37. Morin-Papunen L, Vauhkonen I, Koivunen R, Ruokonen A, Martikainen H, acteristics of the metabolic syndrome in women with polycystic ovary syn- Tapanainen JS 2003 Metformin versus ethinyl estradiol-cyproterone acetate in drome. J Clin Endocrinol Metab 90:1929–1935 the treatment of nonobese women with polycystic ovary syndrome: a ran- 20. Ehrmann DA, Liljenquist DR, Kasza K, Azziz R, Legro RS, Ghazzi MN 2006 domized study. J Clin Endocrinol Metab 88:148–156 Prevalence and predictors of the metabolic syndrome in women with poly- 38. Gordon DJ, Rifkind BM 1989 High-density lipoprotein–the clinical implica- cystic ovary syndrome. J Clin Endocrinol Metab 91:48–52 tions of recent studies. N Engl J Med 321:1311–1316 21. Norman RJ, Davies MJ, Lord J, Moran LJ 2002 The role of lifestyle modifi- 39. Harborne L, Fleming R, Lyall H, Sattar N, Norman J 2003 Metformin or cation in polycystic ovary syndrome. Trends Endocrinol Metab 13:251–257 antiandrogen in the treatment of hirsutism in polycystic ovary syndrome. 22. Ehrmann DA 2005 Polycystic ovary syndrome. N Engl J Med 352:1223–1236 J Clin Endocrinol Metab 88:4116–4123 23. Diamanti-Kandarakis E, Baillargeon J-P, Iuorno MJ, Jakubowicz DJ, Nestler 40. Rautio K, Tapanainen JS, Ruokonen A, Morin-Papunen LC 2005 Effects of JE 2003 A modern medical quandary: polycystic ovary syndrome, insulin metformin and ethinyl estradiol-cyproterone acetate on lipid levels in obese resistance, and oral contraceptive pills. J Clin Endocrinol Metab 88:1927–1932 and non-obese women with polycystic ovary syndrome. Eur J Endocrinol 24. Costello M, Shrestha B, Eden J, Sjoblom P, Johnson N 2007 Insulin-sensitising 152:269–275 drugs versus the combined oral contraceptive pill for hirsutism, acne and risk 41. Tikkanen MJ, Nikkila EA 1987 Regulation of hepatic lipase and serum li- of diabetes, cardiovascular disease, and endometrial cancer in polycystic ovary poproteins by sex steroids. Am Heart J 113:562–567 syndrome. Cochrane Database Syst Rev:CD005552 42. Chasan-Taber L, Stampfer MJ 1998 Epidemiology of oral contraceptives and 25. Zawadzki JK, Dunaif A 1992 Diagnostic criteria for polycystic ovary syn- cardiovascular disease. Ann Intern Med 128:467–477 drome: towards a rational approach. In: Dunaif A, Givens JR, Haseltine FP, 43. Baillargeon J-P, McCliah DK, Essah PA, Nestler JE 2005 Association between Merriam GR, eds. Polycystic ovary syndrome. Boston: Blackwell Scientific the current use of low-dose oral contraceptives and cardiovascular arterial Publications; 377–384 disease: a meta-analysis. J Clin Endocrinol Metab 90:3863–3870 26. Hatch R, Rosenfield RL, Kim MH, Tredway D 1981 Hirsutism: implications, 44. Villaseca P, Hormaza P, Cardenas I, Oestreicher E, Arteaga E 2004 Ethinyl etiology, and management. Am J Obstet Gynecol 140:815–830 estradiol/cyproterone acetate in polycystic ovary syndrome: lipid and carbo- 27. Goodman A 1999 Terminology and differential diagnosis of genital tract hydrate changes. Eur J Contracept Reprod Health Care 9:155–165 bleeding in women. In: Snyder PJ, Utiger RD, eds. Up to date in endocrinology 45. Lord JM, Flight IH, Norman RJ 2003 Insulin-sensitising drugs (metformin, and diabetes. Wellesley, MA: UpToDate, Inc. troglitazone, rosiglitazone, pioglitazone, D-chiro-inositol) for polycystic ovary 28. Escobar-Morreale HF, Serrano-Gotarredona J, Varela C, Garcia-Robles R, syndrome. Cochrane Database Syst Rev:CD003053 Sancho JM 1997 Circulating leptin concentrations in women with hirsutism. 46. Harborne L, Fleming R, Lyall H, Norman J, Sattar N 2003 Descriptive review Fertil Steril 68:898–906 of the evidence for the use of metformin in polycystic ovary syndrome. Lancet 29. Escobar-Morreale HF, Lasuncio´n MA, Sancho J 2000 Treatment of hirsutism 361:1894–1901 with ethinyl estradiol-desogestrel contraceptive pills has beneficial effects on 47. Azziz R, Sa´nchez LA, Knochenhauer ES, Moran C, Lazenby J, Stephens KC, the lipid profile and improves insulin sensitivity. Fertil Steril 74:816–819 Taylor K, Boots LR 2004 Androgen excess in women: experience with over 30. Vermeulen A, Verdonck L, Kaufman JM 1999 A critical evaluation of simple 1000 consecutive patients. J Clin Endocrinol Metab 89:453–462

JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.